Agentic Design Patterns: A System-Theoretic Framework

TL;DR

Problem: foundation model based agentic AI suffers from hallucinations and brittle reasoning due to ad-hoc design. Approach: introduce a principled, system‑theoretic framework that decomposes agents into five subsystems and derives 12 reusable design patterns anchored to inter‑subsystem interactions, plus a qualitative ReAct case study. Contributions: a five‑subsystem architecture, a catalog of 12 agentic design patterns with explicit intent, problem, and solution, and a qualitative ReAct analysis diagnosing weaknesses and prescribing improvements. Significance: provides a common vocabulary and practical methodology for building modular, verifiable, and trustworthy autonomous agents, aligning software design patterns with AI agent design.

Abstract

With the development of foundation model (FM), agentic AI systems are getting more attention, yet their inherent issues like hallucination and poor reasoning, coupled with the frequent ad-hoc nature of system design, lead to unreliable and brittle applications. Existing efforts to characterise agentic design patterns often lack a rigorous systems-theoretic foundation, resulting in high-level or convenience-based taxonomies that are difficult to implement. This paper addresses this gap by introducing a principled methodology for engineering robust AI agents. We propose two primary contributions: first, a novel system-theoretic framework that deconstructs an agentic AI system into five core, interacting functional subsystems: Reasoning & World Model, Perception & Grounding, Action Execution, Learning & Adaptation, and Inter-Agent Communication. Second, derived from this architecture and directly mapped to a comprehensive taxonomy of agentic challenges, we present a collection of 12 agentic design patterns. These patterns - categorised as Foundational, Cognitive & Decisional, Execution & Interaction, and Adaptive & Learning - offer reusable, structural solutions to recurring problems in agent design. The utility of the framework is demonstrated by a case study on the ReAct framework, showing how the proposed patterns can rectify systemic architectural deficiencies. This work provides a foundational language and a structured methodology to standardise agentic design communication among researchers and engineers, leading to more modular, understandable, and reliable autonomous systems.

Figures (4)

• Figure 1: A system-theoretic agent architecture. The model illustrates the internal structure of an agent as nested functional layers, comprising four core subsystems and one extensible subsystem (IAC, highlighted as optional).
• Figure 2: The agent's cognitive cycle. This diagram illustrates the dynamic interaction flows between the four core subsystems and the optional communication subsystem (IAC, shown with dashed lines).
• Figure 3: Conceptual Sankey diagram illustrating the relationships between identified classes of problems, core agent subsystems, and the 12 proposed agentic design patterns. Flow widths indicate qualitative relevance.
• Figure 4: A conceptual diagram showing how ReAct can be enhanced by integrating our proposed agentic design patterns.